The methylation status of DNA influences many biological processes during mammalian development and is known to be highly aberrant in cancer. In mammalian cells, DNA methylation occurs primarily as symmetrical methylation of cytosine in the context of the dinucleotide CpG, and the presence of high levels of 5-methyl- cytosine (5mC) at promoters is generally correlated with diminished gene expression. We recently discovered that the TET proteins TET1, TET2 and TET3 constitute a new family of dioxygenases that utilize molecular oxygen and the cofactors Fe(II) and 2-oxoglutarate to oxidize 5mC to 5-hydroxymethylcytosine (5hmC) in DNA. As a result, TET proteins alter DNA methylation status in a novel and hitherto unprecedented way. Simultaneously, several labs reported that TET2 mutations are frequently associated with myelodysplastic syndromes (MDS), myeloproliferative neoplasms (MPN) and myeloid malignancies such as chronic myelomonocytic leukemia (CMML) and acute myeloid leukemia (AML). Two other genes frequently mutated in these patients include those encoding the DNA methyltransferase DNMT3A and the polycomb group protein ASXL1, a component of a complex that deubiquitinates H2A. In this proposal we will explore, at a molecular level using appropriate mouse models, the roles of Tet2 and Tet3 in hematopoiesis and myeloid function (Aim 1). In Aims 2 and 3, we will investigate the relation of Tet2/ Tet3 to Dnmt3a and Asxl1 respectively. We have developed many methods and generated many reagents relevant to these proposed studies, including quantitative methods to measure overall genomic levels of 5hmC in bone marrow samples from patients with MDS/ MPN/ CMML/ secondary AML; and innovative strategies for mapping the genomic location of 5hmC and profiling 5hmC at single-base resolution. We have generated mice with conditional disruption of the Tet2 and Tet3 genes; and have uncovered a novel relation between 5hmC and the polycomb complex. WE have shown in ES cells that 5hmC is present predominantly at the promoters of genes that are (i) inactive but poised to be expressed upon ES cell differentiation; (ii) bear dual (bivalent) H3K4me3 and H3K27me3 marks; and (iii) are bound by components of the polycomb complexes PRC1 and PRC2 (the H3K27me3 mark is deposited by PRC2). By defining the genome-wide changes in DNA methylation, DNA hydroxymethylation and histone modifications that occur as a result of loss of function of Tet2 and Tet3 and selected leukemia-associated mutations in Dnmt3a and Asxl1, our proposed studies will provide fundamental insights into how these proteins control chromatin structure and the epigenetic landscape at their target genes. This information will help us understand how changes that occur as a result of somatic mutations in these gene products in stem cells might predispose to myeloid cancers in humans.